Skip to content

Latest commit

 

History

History
151 lines (88 loc) · 3.75 KB

README.template.md

File metadata and controls

151 lines (88 loc) · 3.75 KB

Sphinx Search docker file

Version: %%SPHINX_VERSION%%

You can read the official changelog for details on changes.

Content

A Sphinx Search instance builded from source.

Supports:

  • stemming (via libstemmer, link)

  • xml (with expat and iconv)

  • postgresql

  • mysql

  • odbc

  • regular expression filter (via RE2 engine, version %%GOOGLE_RE_VERSION%%, link)

  • lemmatization

    • /var/diz/sphinx/ru.pak (russian dict)
    • /var/diz/sphinx/en.pak (english dict)
    • /var/diz/sphinx/de.pak (deutsch dict)

Exposed ports

  • 9312 for client connections

  • 9306 for SQL connections

Mount points

This image provides some directories for configurations, logs and other files:

  • /var/idx/sphinx

  • /var/log/sphinx

  • /var/lib/sphinx

  • /var/run/sphinx

  • /var/diz/sphinx

All this directories are data volumes.

Scripts

The available scripts are:

  • searchd.sh, to start searchd in the foreground (needed also for real-time indexes)
  • indexall.sh, to index all the plain indexes (i.e., indexer --all) defined in the configuration

Path

Both scripts and Sphinx Search's tools (e.g., the spelldump tool) are available from the PATH.

To list all Sphinx Search's tool you can execute:

docker run -it leodido/sphinxsearch:%%SPHINX_VERSION%% ls /usr/local/bin
# indexer  indextool  searchd  spelldump	wordbreaker

Installation

You can clone this repository and manually build it.

cd dockerfiles/sphinxsearch\:%%SPHINX_VERSION%%
docker build -t leodido/sphinxsearch:%%SPHINX_VERSION%% .

Otherwise you can pull this image from docker index.

docker pull leodido/sphinxsearch:%%SPHINX_VERSION%%

Usage

The simplest use case is to start a Sphinx Search container, attach to it and do whatever you want with it:

docker run -i -t leodido/sphinxsearch:%%SPHINX_VERSION%% /bin/bash

Daemonized usage (1)

Assume that we want to index our documents into some real-time indexes.

Given a Sphinx Search configuration file (e.g., sphinx.conf) in our current directory (i.e., $PWD), we have to share its content with the container using docker option -v.

We also want to link to exposed 9306 port to query Sphinx Search from the host machine.

So, the command to run a daemonized instance of this container is:

SS=$(docker run -i -t -v $PWD:/usr/local/etc -p 9306 -d leodido/sphinxsearch:%%SPHINX_VERSION%% searchd.sh)

Now we want to see to which host address it has been linked:

docker port $SS 9306 # which returns 49174

And eventually try to connect to it:

mysql -h 0.0.0.0 -P 49174

We can now index documents into our Sphinx Search container or perform queries against it.

Daemonized usage (2)

Assume that we want to index our documents into some plain indexes.

We need:

  1. the data source files (e.g. XML files structured as demanded by the Sphinx Search's xmlpipe2 driver)

  2. a valid Sphinx Search configuration file that defines our plain indexes and their sources

  3. a way to querying Sphinx Search from the host machine (e.g., using IP 127.0.0.1 and port 9306)

So, assuming that in our current directory (i.e., $PWD) we have these files, we run a daemonized instance of Sphinx Search as follow:

docker run -i -t -v $PWD:/usr/local/etc -p 127.0.0.1:9306:9306 -d leodido/sphinxsearch:%%SPHINX_VERSION%% indexall.sh

This way we have indexed our documents and started serving queries.

Again, if you want to query from the host machine:

mysql -h 127.0.0.1 -P 9306

Analytics